and Reptdsim in the Lunar Rai/s, &c. Mf7 



specific lightness, and the discs must be kept perfectly free from 

 damp. Care must be taken, also, that the capsule is fairly 

 placed on the pivot, which ought to be very fine. Every source 

 of partial light and heat ought, as far as possible, to be exclud- 

 ed. And the instruments must be kept covered from the light 

 some hours before they are used, as they will not move to a 

 subdued degree of light if they have been exposed to a greater. 

 Their sensibilities are considerably blunted for a time, if expo- 

 sed to powerful light. We must also keep at some distance 

 from the instrument when making the trials, as the heat and 

 electricity that escape from our bodies are a source of attrac- 

 tion. All these bodies are much influenced by the solar beam. 

 But nothing yet observed, if used in eqical weights, moves so re- 

 gularly as the magnetic steel to the sun's influence, which is af- 

 fected in a way peculiar to itself. 



Two causes are assigned for the phenomenon, that all bodies 

 of sufficient specific lightness, having two flat sides approaching 

 to planes, and free to move, turn always the edges of their planes 

 to the source of light, and their planes parallel to the line of in- 

 cidence. One of the causes appears to be a sort of elective at- 

 traction, which lightj like electricity, has for the points or edges of 

 bodies. The other reason is, that all bodies kept excluded from 

 light, are, when exposed to it, first attracted by it ; and when, 

 from their colour or opacity, they have absorbed a certain quan- 

 tity of the rays, are then repelled by it. The rays of the 

 sun evidently soon repel all the substances mentioned ; and 

 when they turn their edges, they are in that position where they 

 receive the least jx)ssible impulse from the rays. As a vane is 

 turned by the mechanical force of a current of air, these instru- 

 ments are turned by the repelling power of the beams of light. 

 Bodies, quite transparent, are not taken into the account. The 

 silver leaf is a half exception to this general law, but it is al- 

 most colourless and pohshed, and therefore absorbing but a small 

 portion of the light, and quickly parting with it ; it assumes d 

 vibratory motion, first turning its plane, and then its edge, to a 

 strong light, and thus continually moving in the arc of a qua- 

 drant. To a feeble light it stands with its plane confronting it. 



The motion of the leathers seems chiefly to be occasioned by 

 attraction. And as each filament of a peacock'*s feather, of the 

 size used, has about 4000 piles upon it, each disc, at a moderate 



